The one sign that settles it
Most signs a product is overpriced are reasons to look closer. Only one of them is a finding: today's price sitting well above the band this product has traded in for months, with nothing to explain the gap. Everything else on this page is a prompt to run that check.
A high price and an overpriced one are different things. A product can cost three times what a competitor charges and be fairly priced, if the extra is buying materials, a warranty someone honours, or a machine that lasts twice as long. Overpriced means paying for something that is not there. Whether the product is worth owning at any price is the pillar this sits under, deciding whether something is worth it. The price half of the question, and the full method for judging one against its own history, is set out in whether this is a good deal.
A reference price nobody pays
The most common source of an inflated price is not the price at all. It is the number beside it.
Retailers measure a discount against a reference they choose, and nothing obliges that reference to be a price the product sold at in volume. It can be a manufacturer's suggested figure set years ago and never revised, a price the listing carried for one week before a promotion began, or simply the highest number the page has ever displayed.
Once that reference exists, every price underneath it looks like a saving, permanently and automatically. A $50 offer with a $200 reference shows a 75% badge every day of the year. The arithmetic is correct. The saving is close to nothing if the product has spent months trading near $52.
The three numbers that get used interchangeably here are worth separating, because the confusion starts with them: what a manufacturer suggests, what one retailer displays as standard, and what the thing actually costs across sellers right now. Only the third is transactable. The difference between them is covered in MSRP against street price.
A sale that never ends
If a product is on sale every time you look at it, the sale is not an event. It is the price.
This is the easiest signal to check and the one people most often dismiss, because a permanent discount does not feel like a trick. Nobody is lying. The listing genuinely is cheaper than its reference, and it always will be, which is precisely what makes the percentage meaningless as a signal about timing.
The related pattern is the hill rather than the cliff: a price that drifts upward for two or three weeks and then falls back to roughly where it started, with a large percentage attached to the fall. On a chart it is unmistakable once you have seen one. On the listing it is invisible.
A brand premium buying nothing
Brand premiums are frequently legitimate. Better materials cost more. A warranty someone actually honours costs more. Support that answers the phone costs more, and so does designing something properly rather than ordering it from a catalogue.
The premium becomes a markup when none of those apply. A large share of the online catalogue is the same factory product sold under many different names, each with its own invented brand, its own logo composited into the same photography, and its own independent set of reviews. Five different products at five prices can be one product.
The tells are recognisable. An unpronounceable brand name with no website. Specification tables matching word for word across listings, including the same errors. A brand whose catalogue spans a kettle, a set of resistance bands and a car phone mount. None of that makes the product bad, and for undemanding categories the factory output at the lower price is a genuinely good trade. What it does mean is that the higher-priced identical listing is charging for a name with nothing behind it.
A bundle padded with list prices
Bundles are frequently cheaper than their components and are also the easiest place to manufacture an apparent saving.
The method is simple: include something with a high stated list price and a low real value, then present the sum of the list prices as what the bundle is worth. Cables, cases, cleaning kits, software licences and extended-warranty vouchers all do this work well, because each has a printed price nobody pays.
The check is to price the components separately at what they actually sell for, not at what the bundle page says they are worth, and compare that total with the bundle price. About as often as not, the individual purchase wins, and you get only the things you wanted.
Multipacks and larger sizes need the same treatment. Convert everything to a price per item, per litre or per kilogram, and the comparison becomes possible. Quite regularly the small size is cheaper per unit, which is the opposite of what the format implies.
A launch price presented as a settled one
New products are priced at a level that will not last, and there is nothing improper about that. Electronics drift below their launch price within months. The outgoing model falls further when the replacement lands, usually by more than any promotional discount it saw during its life.
A launch price is overpriced only in the sense that it is temporary. Paying it because you need the thing now is a legitimate choice. Paying it while believing you found a deal is the failure, and the two feel identical at checkout.
The signal to look for is a listing with a short price history and a large discount badge. A product that has existed for three weeks has no band for the percentage to be measured against, so whatever the badge is comparing to is a number rather than a history.
What a high price legitimately buys
It is worth naming the opposite case, because a page listing only the warning signs teaches people to buy the cheapest thing, which is its own expensive mistake.
In most categories the value curve rises steeply at the bottom and flattens well before the top. The step from the cheapest option to a competent mid-range one is usually large and worth paying for. The step from mid-range to premium is smaller and more specific: better materials, a longer warranty, a feature you may or may not use. Beyond that, price increasingly reflects brand rather than function.
Two things genuinely justify paying more. The first is lifespan, because a product that lasts twice as long at one and a half times the price is cheaper per year. The second is the cost of failure: on anything where a breakdown is expensive, dangerous or simply disruptive, paying for the version that fails less often is not a premium, it is insurance.
There is also a version of this that runs the other way. A price far below the category norm is a signal too, and usually a more urgent one: something has been removed to reach it. Thinner materials, a shorter warranty, a smaller cell, a support address that does not answer. The cheapest option in a category is rarely the best value in it, and the gap between the cheapest and the competent is where most regret is manufactured.
So the honest test is not whether the price is high. It is whether anything on the page explains it, and whether the product's own selling history agrees with the number in front of you.